2 Rules of Thumb When Using Customer Data for Personalization

For the better part of the last decade, marketers have invested and continue to invest in personalization. What was once perceived as an interesting opportunity to improve customer engagement has quickly evolved into an important strategic lever in supporting digital commerce performance.

Gartner’s inaugural digital commerce survey revealed a dominant school of thought: that improving the overall customer experience is key to supporting digital commerce performance. More specifically still, personalization appears to be a key digital commerce enabler within that.

This does makes sense. Personalization is rooted in customer data, and the more customer data can be acquired, the better brands can offer more timely and relevant experiences. This can be done through the delivery of real-time delivery, customized content, and a more connected journey overall.

However, acquiring and using customer data is tough and is likely to get even tougher given growing data complexity, government regulations, and general privacy concerns. Gartner’s recent personalization survey revealed a marked behavioral change, with approximately 74% of customers (both consumers and B2B customers) actively disabling all tracking and muting brand communications.

Rule of thumb #1: Less Data is More

Gartner’s latest personalization study – which included a global survey of ~1500 customers who interacted with a personalized digital experience – identified one approach (‘tailored help’) that was particularly impactful, with a potential uplift of over 30% against Gartner’s commercial benefit index (a measure that tracks brand intent, purchases, repeat purchases and growth).

Tailored help is an approach grounded in helping customers progress in their path to purchase in a brand agnostic fashion. It requires brands to understand their customer needs and deliver an experience that provides easy to use, actionable help. Interestingly, tailored help only requires a very limited amount of customer data to be effective. Why? Two reasons….

The first is that customers value above all messages and content that is intended to be helpful… and brands need little data to create helpful experiences.

The second and more significant reason is because there is a real risk with using too much customer data. That is to say, the more data we leverage for the purpose of delivering more personalized messaging or content, the higher the risk we will deliver experiences that customers will perceive as inappropriate, invasive or (as is more commonly cited) “creepy”. In fact, Gartner research found that the consequence of over-personalizing was significant, with 37.5% of customers more likely to stop doing business with a brand upon receipt of “creepy” communications versus “irrelevant” ones.

Rule of thumb #2: Not All Data Types Are Equal

When it comes to leveraging customer data, not only is it a question of volume but also a question of data type. Across all industries assessed in Gartner’s global personalization survey, customers appear generally comfortable with brands collecting data that relates to their shared experiences. For example, data such as ‘purchase history’ and ‘personal shared preferences’ is seen as appropriate to collect.

However, people are significantly less comfortable when brands leverage data that has no apparent connection to the historical ‘brand – customer’ relationship. For example, data relating to ‘inferred personal preferences’, ‘important life events’ or ‘online browsing history’ are best avoided.

It is worth noting that despite a very similar patterns of results across all industries assessed, the world of B2B does appear to be more comfortable sharing most types of data for the purpose of receiving a more personalized experience overall. However, the risk of getting personalization wrong also appears higher, with a higher proportion of B2B brands indicating that they would stop doing business with brands that poorly and/or over-personalize.